Intraoperative use of indocyanine green fluorescence imaging in rectal cancer surgery: The state of the art

Abstract

Indocyanine green (ICG) fluorescence imaging is widely used in abdominal surgery. The implementation of minimally invasive rectal surgery using new methods like robotics or a transanal approach required improvement of optical systems. In that setting, ICG fluorescence optimizes intraoperative vision of anatomical structures by improving blood and lymphatic flow. The purpose of this review was to summarize all potential applications of this upcoming technology in rectal cancer surgery. Each type of use has been separately addressed and the evidence was investigated. During rectal resection, ICG fluorescence angiography is mainly used to evaluate the perfusion of the colonic stump in order to reduce the risk of anastomotic leaks. In addition, ICG fluorescence imaging allows easy visualization of organs such as the ureter or urethra to protect them from injury. This intraoperative technology is a valuable tool for conducting lymph node dissection along the iliac lymphatic chain or to better identifying the rectal dissection planes when a transanal approach is performed. This is an overview of the applications of ICG fluorescence imaging in current surgical practice and a synthesis of the results obtained from the literature. Although further studies are need to investigate the real clinical benefits, these findings may enhance use of ICG fluorescence in current clinical practice and stimulate future research on new applications.

Key Words: Indocyanine green, Fluorescence imaging, Near infrared, Rectal cancer, Total mesorectal excision, Anastomotic leakage

Core Tip: There is growing interest in real-time fluorescence-guided surgery. The intraoperative use of indocyanine green (ICG) during rectal cancer surgery has found many applications over time. Given the wide availability in current practice, it is important for clinicians to be aware of all potential uses of ICG fluorescence technology in order to facilitate the procedures, limit injuries, and improve outcomes. Herein, we provide a concise overview of the literature regarding the use of ICG fluorescence imaging in this setting.

INTRODUCTION

Colorectal carcinoma is the third most common cancer for both men and women and the second leading cause of cancer-related deaths[1]. Primary rectal localization occurs in 35% of cases[2]. Although multidisciplinary management of rectal cancer is the standard of care, surgery remains the cornerstone of curative treatment. Total mesorectal excision (TME), described for the first time by Heald et al[3,4], should now be performed routinely in cases of middle and lower rectal tumors after neoadjuvant therapy for locally advanced cancer (T3-4 and/or N+)[5]. Respect for the principles of surgical oncology, including complete TME, and negative distal and circumferential resection margins is mandatory to achieve improved survival. Also, much effort has been made to achieve faster and enhanced recovery[6,7] and better quality of life after rectal cancer surgery[8] over time. Advances in oncology research[9] have resulted in increasing roles for the latest technologies in surgery. Enhanced video/camera systems[10], robotic technology[11], powered staplers[12], and specialized operating platforms for natural orifice transluminal endoscopic surgery[13] are just a few examples of surgical innovations in colorectal surgery introduced in the last decades. In this setting, laser fluorescence using indocyanine green (ICG) dye is a promising and widespread real-time technology because of its easy accessibility, accuracy and cost-effectiveness.

The concept of ICG intraoperative angiography is based on the ability of ICG to absorb near-infrared light (NIR) at 800 nm and to emit fluorescence at a wavelength of 830 nm. Albumin is the most important intravascular binding-protein for ICG so that tissue microperfusion is revealed by the presence of fluorescence. In brief, a bolus of ICG is injected into the patient intravenously, NIR light is then absorbed by ICG in the tissue, and the resulting fluorescence is a reflection of perfusion. After the introduction of ICG angiography in clinical practice in 1989 to evaluate choroidal circulation[14], fluorescence imaging technology has been used in hepatobiliary[15,16], gastric cancer[17], gynecologic cancer[18], breast cancer[19] and transplantation surgery[20]. Furthermore, many colorectal surgeons routinely use ICG imaging to assess bowel viability during colorectal anastomosis. However, several applications of ICG angiography have been described in colorectal surgery with increasing interest for rectal cancer resection. In fact, the implementation of minimally invasive rectal surgery using new approaches (e.g., robotic, and transanal) required the improvement of optical systems. In that regard, ICG fluorescence has optimized the intraoperative vision of anatomical structures by the enhancement of blood and lymph flow. The aim of this review is to identify and synthesize data from original research evaluating any possible application of ICG fluorescence imaging in rectal cancer surgery. In the last 5 years, the use of ICG fluorescence in rectal cancer surgery has attracted great interest. The items regarding different intraoperative applications are summarized below.

ASSESSMENT OF VASCULAR PERFUSION AT THE ANASTOMOTIC SITE

Anastomotic leakage (AL) is the most feared complication after TME because it is associated with increased mortality, reoperation, and definitive stoma formation[21,22]. Additionally, the relationship between AL and local recurrence has been found to lead to significant differences in long-term outcomes[23]. It has a significant impact on postoperative functional outcome[24] and increases the economic burden of public health systems[25]. That easily explains the importance of developing prevention strategies in order to reduce the amount of related complications affecting survival and quality of life. The AL rate after anterior rectal resection ranges from 3% to 23%[26,27]. AL is defined by the International Study Group of Rectal cancer as a defect at the level of the anastomotic site that allows communication between the intraluminal and extraluminal compartments identified through clinical evaluation with digital rectal exploration, endoscopic examination, computed tomography radiological evidence of contrast leakage through the suture gap or the presence of perianastomotic hydro-aerial collection. It is also classified as grade A if it does not affect the postoperative course, grade B when conservative management antibiotic therapy or percutaneous/transanal drainage is required, and grace C when surgical revision is required[26,28].

Various risk factors are associated with AL, but insufficient blood perfusion is generally considered as one of the main causes[29,30]. ICG is a water soluble, tricarbocyanine dye that provides real-time visualization of vascular structures by emitting fluorescence when stimulated by polarized light[31]. The evaluation of perfusion of the supposed proximal section line starts within 60 s of an intravenous bolus injection of ICG. An NIR camera detects the fluorescence of the microcirculation of the colonic wall receiving adequate vascularization. Therefore, ICG fluorescence reflects the colonic stump perfusion allowing the choice of the most appropriate site for the final section and then for the anastomosis (Figure 1). A second check can be also performed after the anastomosis construction both by injecting a second bolus of ICG to verify fluorescence of the stumps and by transanal endoscopic suture line visualization. Some limitations of the technique are related to its being, once again, a subjective evaluation by the surgeon of the intensity of the emitted infrared light. Furthermore, it is not a completely standardized technique as there are variables related to the dose of injected ICG (0.013-0.89 mg/kg), the proximity of the laparoscope to the colic wall, the number and type of checks performed, and differences in the equipment available on the market[32-34].

Figure 1 Evaluation of intestinal perfusion after intravenous injection of indocyanine green during total mesorectal excision allows identification of the section line before anastomosis construction (orange arrow).

Kudszus was the first to show the usefulness of ICG fluorescence angiography (FA) in colorectal surgery[35]. The study reported that ICG FA led to a change in the location of the planned proximal resection line in 13.9% of patients. ICG FA significantly reduced AL by 4% compared with the control group. Furthermore, Jafari et al[36] assessed the utility of ICG FA in left colectomy and anterior resection in the PILLAR II study. The incidence of AL in their study was 1.4%, with a change in the surgical strategy in 8% of patients. Boni et al[37] reported that ICG FA could be safely and effectively performed in rectal surgery. The use of IGC-FA changed the surgical plan in 4.7% of their patients and AL did not occur in the ICG group, compared with an incidence of 5.2% in the control group. However, the differences were not statistically significant. The studies agree on the safety and feasibility of the technique and demonstrate its usefulness in the assessment of tissue perfusion.

Several comparative retrospective and prospective studies investigated the relationship between the use of ICG and the 30-d AL rate as the primary outcome[31,33,38-40]. They showed a statistically significant correlation between ICG FA and reduction in the risk of AL associated with the modification of the proximal section line. In a propensity score-matched analysis[41], AL rates of Clavien-Dindo grade ≥ II and ≥ III were 10.4% (22/211) and 9.5% (20/211) in the non-ICG FA group and 4.7% (10/211) and 2.8% (6/211) in the ICG FA group, respectively. Similarly, Foo et al[42] demonstrated that the use of ICG FA was significantly associated with a lower AL rate in TME (4.7% vs 11.6%; P = 0.043) but not non-TME resections. However, not all studies agree on the results. Two propensity score-matched studies failed to demonstrate a statistically significant difference in the rate of AL between the two compared groups, although both reported a change in the proximal section after ICG injection of 27.1% and 18.18%, respectively[43,44]. Most of the studies had limitations such as a small sample size or their retrospective nature. Only two randomized controlled trials (RCTs) were published. In the first one[45], differences in AL or reoperation rates between ICG and the control group (5% vs 9% and 6.7% vs 6.5%) were not significant. Although the authors confirmed the efficacy of ICG in bowel viability assessment during left colectomy or anterior resection, a real advantage related to AL was not demonstrated. The FLAG trial[46] involved 377 patients who underwent anterior rectal resection, 187 in the ICG FA group and 190 in the control group. The results showed that changes in the transection line were performed in almost 20% of patients. A decrease in AL was achieved using ICG FA, but the difference was statistically significant only in cases with low rectal anastomosis (14.4% with ICG FA vs 25.7% without ICG FA; P = 0.04). Two recent meta-analyses assessed the role of ICG FA imaging on the incidence of AL after rectal cancer surgery[34,47]. In a pooled analysis of 2088 patients, the AL rate in the ICG group was significantly lower than that in the control group, and the intraoperative use of ICG was associated with a decreased overall complication rate and reduced reoperation rate. However, both analyses suffer from the same limitations as the studies taken into consideration and the lack of RCTs in the analysis.

Finally, evaluation of bowel perfusion by ICG FA was reported exclusively after transanal total mesorectal excision (TaTME) by Mizrahi et al[48]. In their retrospective cohort of 54 patients who received a very low anastomosis, a check was performed before proximal transection and at the completion of anastomosis. In 18.5% of the cases the surgeon changed the proximal resection margin because of impaired fluorescence. All those anastomoses were shown to be successful at the second control. Two patients (3.7%) suffered from AL, and in neither of them was the splenic flexure mobilized. Furthermore, ICG FA improved clinical outcomes also after robotic sphincter-saving rectal resections[49]. In conclusion, the use of fluorescent angiography with ICG injection has proven to be a safe and feasible method to evaluate bowel perfusion whatever the surgical approach for rectal resection. At the time of the construction of the anastomosis, ICG FA can influence decision making by reconsidering the resection line. Several studies have found a significant decrease in the rate of AL after ICG FA imaging, and that has had a large impact on recovery. Future high quality trials should confirm the impact on AL rates and standardize the technique.

URETHRA VISUALIZATION DURING TRANSANAL TME

TaTME is a relatively new procedure[50] for the curative resection of rectal tumors. It was developed to overcome the difficult dissection at the lower third of the rectum, especially in obese male patients and/or bulky tumors. Some retrospective series have reported that enhanced visualization of the dissection plane allowed better nerve preservation, improved resection margins, and improved functional outcomes compared with laparoscopic TME[51,52]. However, the bottom-up transanal approach is not without complications. Incidence of iatrogenic urethral injuries has been reported, ranging from 1% to 6.7% during TaTME procedures[53]. An international inquiry reported 34 urethral injuries from 32 surgical teams worldwide between 2010 and 2017, resulting in a significant postoperative morbidity rate of 26%[50]. However, there is still concern that urologic injuries during TaTME may be underreported and that their incidence might be related to surgeon experience. Therefore, enhancement of urethral visualization should be considered useful and advantageous in the early learning experience. Several bioimaging modalities exist that can improve urethral identification, including ICG NIR fluorescence[54]. Different systems have been successfully used to detect the urethra by ICG fluorescence imaging such as the IRIS ureteral kit (Stryker, Kalamazoo, MI, United States)[55] or the PINPOINT laparoscopic system (Stryker, United States)[56] with intraurethral ICG injection or infiltration adjacent to the catheter in the urethra, respectively. Experimental studies demonstrated that direct ICG instillation into the urethra or through a urine catheter for NIR fluorescence imaging seem to be easily applicable and clinically reproducible during TaTME[55-57]. Although an open-label clinical feasibility study (NCT03204201) with intraoperative direct instillation of ICG into the urethra for low rectal cancers was terminated because of technique failures, implementation of the technique has been described by some authors. Barnes et al[58] evaluated the efficacy of two novel methods in cadaveric models. In the first, ICG mixed with silicone was infiltrated into 10-Fr one-way Foley catheter and allowed to set for 1 wk. In the second, new preclinical IRDye 800BK (LI-COR Biosciences^®, Lincoln, Nebraska, United States) was infiltrated directly into the urethra via the urethral meatus prior to dissection. Both methods were effective in identifying the fluorescence located only within the urethra. IRDye 800BK provided a greater depth of penetration than the ICG-silicone mix, suggesting it could be a more satisfactory alternative to ICG. In addition, Barberio et al[59] demonstrated the superior brightness of near-infrared coating of equipment (NICE) coated catheter compared with ICG-based solutions in cadaveric experiments by exhibiting a higher fluorescence intensity than urinary catheters filled with ICG. In conclusion, no final specific recommendations can be drawn from the clinical use of ICG fluorescence imaging to identify and prevent urethral injuries during TaTME procedures because of the very limited data available. However, future studies will have to take into account that fluorescence technology plays a major role in this setting.

URETER IDENTIFICATION

The incidence of iatrogenic urethral injuries (IUI) ranges from 0.24% to 1.95% in colorectal surgery, and rectal cancer is considered a risk factor for IUI because of the close proximity of the ureters to the dissection plane[60], similar to the risk with deep pelvic endometriosis[61,62]. Despite its low incidence, IUI significantly affects postoperative morbidity, mortality, length of stay and hospital charges[60]. Visualization of the ureters is thus advocated during pelvic surgery by the visible peristalsis that occurs when the ureter is gently pressed (Kelly’s sign). However, adhesions, obesity, and an incorrect plane of dissection contribute to the lack of or incorrect recognition of the ureter, which can jeopardize its integrity. For that reason, a selective use of prophylactic urethral stents in high risk procedures is commonly accepted, but there is no sufficient evidence to support a decrease in IUI or intraoperative identification[63,64]. In that setting, interest in fluorescence imaging has been increasing over time. While contrasting results were found for intravenous administration of methylene blue dye to urethral detection in colorectal surgery[65,66], ICG FA proved to be a viable alternative to real-time ureter identification and IUI prevention. Before surgery, a 6-Fr catheter is placed into the urethral orifice by cystoscopy. As ICG binds to the proteins of the ureteric epithelium[67], a retrograde injection of 5 mg ICG diluted in 2 mL of distilled water is made, and infrared emission is captured by the filtered lens system and electronically converted into green color visualizing ureter location. The technique has proven to be safe and helpful to identify the ureter in several small case-series who underwent minimally invasive pelvic surgery[68-72]. As a catheter insertion of only 1 cm is required, there is a lower risk of IUI during catheterization than during conventional endoscopic stenting procedures, thus avoiding additional cystoscopies to remove the catheter. Furthermore, ICG urethral instillation is less expensive than other fluorescence-based systems such as illuminated catheters[68].

White et al[73] recently evaluated the safety and efficacy of intraurethral ICG FA along with any potential benefit related to the technique during colorectal robotic surgery. In their experience involving 16 patients, there were short procedure times, low morbidity, and reliable urethral identification and avoidance. The United States Food and Drug Administration approval of ICG is limited to intravenous use[74].Therefore, disclosure of intraurethral off-label use would be needed. In contrast, new intravenous fluorescent dyes with renal clearance, such as fluorescein sodium[75] and IRDye^® 800-BK[76] have been used in experimental models to test the penetration of fluorescence in the ureters, with promising results for surgical practice. Additionally, the formulation of ICG in a liposome-based delivery system allows its excretion in urine in animal models and seems a promising fluorophore solution[77,78]. In conclusion, evidence supporting the use of intraurethral ICG instillation in order to improve intraoperative ureter detection is based on few noncomparative feasibility studies involving mixed pelvic surgeries. Despite the efficacy demonstrated in ordinary or complex situations, no study exclusively focused on rectal cancer resections exists to date. It remains to be proven whether this innovation significantly affects surgical procedures and provides clinical benefits by reducing IUI.

LYMPH NODE MAPPING

Lateral pelvic lymph node dissection (LLND) allows the removal of the nodal compartment along the common iliac, internal iliac, and obturator arteries. The lymphatic stations are considered a major cause of locoregional recurrence in rectal cancer and are treated with preoperative chemoradiotherapy and curative resection[79]. While it is widely accepted to perform LLND in selected patients with rectal cancer and lateral lymph nodes that are clinically positive[5], the Japanese Society for Cancer of the Colon and Rectum guidelines[80] recommend LLND even when lateral lymph node metastasis is not detected by preoperative or intraoperative diagnosis. Indeed, LLND is associated with a lower rate of local recurrence compared with TME alone despite no significant differences in either overall survival or local recurrence-free survival[81].

As ICG fluorescence imaging has proven to be a useful tool for identifying lymphatic drainage in colorectal surgery[82,83], ICG-enhanced NIR fluorescence-guided imaging has been used to improve the accuracy and the completeness of LLND. In such cases, ICG fluorescence imaging is carried out the injection of ICG dye into the submucosal layer on the distal side of the tumor through the anus immediately before surgery. In a comparative retrospective series of 42 mid and low rectal cancer patients, the ICG group experienced a significantly lower intraoperative blood loss and a larger number of harvested lateral pelvic lymph nodes[84]. The use of ICG may improve the safety of LLND that is affected by the technical difficulties of the procedure, complicated pelvic wall anatomy, and the effects of preoperative radiation on the tissues. In that setting, real-time identification of lateral pelvic nodes could help to distinguish lymphatic tissue from vascular and nervous structures, thus avoiding postoperative genitourinary dysfunction and providing better surgical staging. However, evidence is still limited[85-87] and additional studies are needed to address the real clinical advantages and standardization of this technique.

A sentinel node (SN) is defined as the first node in the regional peritumoral area that drains the tumor. SN biopsy, in addition to conventional resection, may add clinically significant prognostic information in colorectal surgery[88-90]. NIR laparoscopy with ICG mapping allowed easy intraoperative identification of mesocolic lymphatic drainage and SN during colorectal oncologic resections[91]. Similarly Noura et al[92] described the detection of SN by ICG with an NIR system in 25 patients who had no preoperative diagnosis of metastatic lateral pelvic lymph nodes. The success rate of detecting the lateral SN was 92%, and 100% concordance was observed between SN and dissected lateral lymph nodes status. That preliminary study highlighted the feasibility and reliability of lateral SN biopsy as a potential discriminator to perform LLND, but the sensitivity may be compromised by preoperative neoadjuvant chemoradiotherapy[93,94].

TUMOR LOCALIZATION

Several reports have described the intraoperative identification of colonic tumors by NIR with ICG fluorescence imaging, with satisfactory results[95-98]. Accurate identification of the location of colorectal tumors is crucial in minimally invasive surgery because of the lack of tactile perception, especially for cancer at an early stage because of its small size or location on a movable part of the colon. As for rectal cancer, precise tumor site localization allows achieving a clear and safe distal resection margin, which may affect not only oncological outcomes but also bowel function and quality of life. In that setting, endoscopic tattooing of rectal tumors, both with a high-definition fluorescence imaging system (Karl Storz GmbH & Co. KG, Tuttlingen, Germany)[93] and the PINPOINT^® endoscopic fluorescence imaging system (PINPOINT system; Novadaq Technologies Inc., Mississauga, ON, Canada)[99], is feasible and has clinical advantages. In a comparative retrospective series, 342 patients scheduled for laparoscopic colorectal resection were enrolled after propensity score matching[100]. The tumor was tattooed in 114 patients. In a subgroup analysis of 160 patients who underwent anterior resection, the tattooed group had a significantly shorter operative time (unlike right and left colectomy), less blood loss, and a shorter hospital stay than the non-tattooed group. In addition, Goo et al[101] compared 200 tattooed colorectal cancer patients (44 rectal cancers) with 879 non-tattooed patients (300 rectal cancers) to evaluate the effect of preoperative colonoscopic tattooing with ICG on adequate lymph node harvest in colorectal cancer. They found that preoperative tattooing in T1 colorectal cancer significantly improved adequate lymph node harvest, with a higher number of retrieved lymph nodes in rectal cancer than in colon cancer.

Fluorescence technology to localize rectal tumors has been developed not only in the field of imaging systems[99], but also by ICG formulation. Fenestrated peritumoral capillaries and impaired lymphatic drainage delay the washout of large molecules from tumors, which has been described as the enhanced permeability and retention effect[102]. The formulation of ICG as a liposome-based delivery system improved tumor-specific localization in experimental models, with the advantages of intravenous injection and better results than free ICG[103,104]. Finally, there are limited data on the role of ICG in the detection of peritoneal carcinomatosis of colorectal origin. Cytoreductive surgery followed by hyperthermic intraperitoneal chemotherapy is the only potentially curative option in patients with limited peritoneal metastases[105]. Intraoperative injection of ICG seems a useful tool to identify peritoneal metastases and detect additional subclinical malignant peritoneal nodules, resulting in modification of the planned surgery in 29% of patients[106]. However, further investigations are required to draw firm conclusions.

OTHER USES

Peritumoral injection of ICG may help the surgeon to perform an adequate dissection along the embryological surgical planes and visualize the relationship with surrounding structures during TaTME[107], conventional laparoscopic TME[108], and abdominoperineal resection (APR)[109]. Omentoplasty is a well-known method to fill the pelvic cavity after APR or in case of complications after rectal cancer surgery. A pilot study of the intraoperative value of NIR fluorescence imaging with ICG to assess omental perfusion after the creation of a pedicled omentoplasty found that a change in decision making occurred in 80% of the cases[110], and a positive impact on the nonhealing rates of patients undergoing salvage surgery for chronic pelvic sepsis was also observed[111].

CONCLUSION

The adoption of ICG fluorescence imaging in rectal cancer surgery and its multiple applications has increased over time. The major field of application is the evaluation of bowel perfusion at the time of anastomosis construction along with a better intraoperative identification of anatomical structures such as the ureter, urethra, lymph nodes, and tumor location. These objectives are relevant because they aim to improve patient safety by avoiding or reducing the risk of complications. However, further investigations are needed to assess the impact of intraoperative ICG fluorescence imaging on clinical, oncological and cost-effective outcomes.